Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    Magic folders

    1

    Event-driven macOS Desktop folders that automatically convert, transcribe, and clean media files using MLX and FFmpeg.

    $2

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned

    Magic folders

    Example session with this skill installed

    Instala as pastas mágicas / Install magic folders

    • Read your context and instructions
    • Compiled the magic folders

    I will set up the Magic Folders automation system on your Mac. This involves verifying your arm64 environment, installing dependencies, and deploying the twelve WatchPath agents to your Desktop.

    Installation Log & System Status

    ActionStatusDetails
    System CheckPASSarm64 detected, macOS 14.5, all binaries present
    DependenciesINSTALLEDffmpeg, mlx-whisper, and pillow confirmed
    RepositoryCLONEDSource moved to ~/Desktop/VIBECODING/pessoal/pastas-magicas
    LauncherCOMPILEDPastasMagicas.app created and signed
    AgentsLOADED12 launchd agents active via WatchPaths
    Folder ScaffoldingCREATED12 folders added to Desktop (2 tagged red for API)

    Active Automation Folders

    ~/Desktop/
    ├── audio>>limpo/          (Voice isolation - Needs ElevenLabs)
    ├── video>>audio/          (MP4 to MP3 conversion)
    ├── video>>legenda/        (Auto-transcription & hardcoded subs)
    ├── video>>melhorestrechos/ (AI clip selection - Needs Gemini)
    ├── video>>gif/            (Video to GIF conversion)
    ├── audio>>transcrever/    (Speech-to-text via MLX)
    └── [6 other specialized processing folders]
    

    Next steps
    1.

    Permissions: Go to Settings > Privacy > Full Disk Access and add ~/Library/Application Support/pastas-magicas/PastasMagicas.app.
    2.

    API Keys: Run open -e ~/Library/Application\ Support/pastas-magicas/.env to add your Gemini and ElevenLabs keys.
    3.

    Test: Drop a short video into video>>audio and run tail -f ~/Library/Logs/pastas-magicas.log to watch the conversion.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    Manually processing media files is a waste of developer time. This skill automates local media conversion, transcription, and cleanup by installing twelve "magic folders" on your macOS Desktop that trigger specific actions the moment a file is dropped in. It uses launchd and WatchPaths to ensure no persistent background processes consume resources when idle.

    What it does

    • Media Conversion converts video to audio, GIF, or MP4 formats automatically.
    • AI Transcription generates text transcripts and burns hardcoded subtitles using mlx-whisper.
    • Audio Optimization performs silence removal, LUFS volume leveling, and AI voice isolation.
    • Content Selection identifies best clips and performs script-based cuts using LLMs.
    • System Monitoring manages background agents via launchctl for event-driven execution.

    How it works

    1. Pre-flight Check verifies system architecture, macOS version, and required binaries like ffmpeg and clang.
    2. Environment Setup installs Python dependencies including mlx-whisper and pillow.
    3. Automated Deployment clones the private repository and runs instalar.sh to compile the launcher and load agents.
    4. Configuration guides the user to add API keys for ElevenLabs and Gemini to the local .env file for advanced modes.
    5. Validation runs test files through the folders and monitors real-time logs to confirm successful processing.

    Frameworks & tools

    This tool is built for macOS (Apple Silicon arm64) using Python 3, MLX, FFmpeg, and launchd. It integrates with ElevenLabs for audio cleanup and Google Gemini for content analysis.

    Why this beats prompting it yourself

    Writing reliable folder watchers that handle file locking, temp directories, and launchd lifecycle management is tedious. This skill handles the idiosyncratic macOS permissions (TCC) and creates a production-ready, event-driven local pipeline without manual CLI execution.

    Use cases

    • Content Creation drop raw footage to get instant social media clips and burned-in captions.
    • Meeting Archiving convert video calls to transcripts and normalized audio files automatically.
    • Workflow Automation migrate automation scripts between Mac workstations during hardware upgrades.
    • Podcast Production clean audio levels and remove silence from raw recordings by moving files.

    Known limitations

    Strictly requires macOS 12+ on Apple Silicon (arm64). Certain AI features require external API keys for ElevenLabs and Gemini. Intel Macs only support eight of the twelve automation modes.

    Requires macOS 12+ and Apple Silicon (arm64) for full functionality. Intel Macs lose transcription and AI clip selection unless modified. Requires manual TCC Full Disk Access for the launcher app.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 8 days ago

    Requires

    Git

    Permissions

    Terminal / Shell
    Network Access
    Read Files
    Write Files
    Listed8 days ago

    What's inside

    Frequently Asked Questions

    This skill creates 12 specialized folders on your macOS Desktop that automatically process any file dropped into them—performing tasks like AI transcription, video-to-GIF conversion, and audio leveling without you ever opening a CLI or media editor.

    It is optimized for Apple Silicon (M1/M2/M3) running macOS 12+. While Intel Macs can run the basic FFmpeg conversions, the advanced AI transcription and MLX features require the Unified Memory architecture of arm64 chips.

    The purchase includes the complete installation script, the local Python environment setup, and the proprietary launchd configuration files that manage the event-driven background agents.

    No, the skill uses event-driven 'WatchPaths,' meaning the automation only consumes CPU/RAM the moment a file is detected; it remains completely idle and resource-neutral when you aren't using the folders.

    While core conversions and MLX transcription happen locally, advanced features like voice isolation and content analysis require your own API keys for ElevenLabs and Google Gemini, which you can easily add to the provided .env file.

    Initial setup is handled by an automated 'instalar.sh' script that checks your system for FFmpeg and Python dependencies, compiles the necessary launcher, and registers the folders with your system's background manager.

    New in Workflow Automation